Re: poor relative G4 performance iBook/PowerBook
One other thought- to get an idea of real speeds you might try copying one of your entire source trees with cp -r dir1 dir2 Or something like that On Wed, Dec 9, 2020 at 6:30 AM Riccardo Mottola wrote: > Hi, > > > Brian Morris wrote: > > % time cat huge_file_copy > > > > Install smartctl to check drive health > > > > I also have a PowerBook 1ghz which is quite a bit faster than my > > 1.5ghz tho could be because the former has a bit more ram and a newer > > hard drive and the latter has a defective 2nd ram slot > > I wanted to investigate this further - except smartctl, I did more > extensive checks with file copying which are interesting, also, now that > keyboard works again, I investigated better what dmesg says. > Here some results: > > dd if=/dev/zero of=output.dat bs=1M count=128 > > PB: 158 MB/s > iBook: 117 MB/s > > dd if=/dev/zero of=output.dat bs=1M count=256 > > PB: 9.4 28.5 MB/s > iBook: 4.8s 56 MB/s > > Copy of 256MB file > PB: 5.18 s > iBook: 2.7 s > > This means that for 128MB there is some "cache effect" and there the > PowerBook beats the iBook. But with 256MB the iBook is almost twice as > fast! 28.5 MB/s is less than UDMA/33... quite slow... > Both the iBook and PowerBook still have PATA of course, since no serial > ata devices existed back then > > the iBook has > [3.774751] pata-pci-macio 0002:20:0d.0: Activating pata-macio > chipset UniNorth ATA-6, Apple bus ID 3 > [3.775921] scsi host0: pata_macio > [4.802731] pata-macio 0.0002:ata-3: Activating pata-macio > chipset KeyLargo ATA-3, Apple bus ID 0 > [4.810545] scsi host1: pata_macio > > I bet the first one is the Hard Disk, the second one is for the optical > drive. > > The PowerBook appears also to have the same UniNorth and KeyLargo... so > I doubt there is a driver issue? Apple reused the same chips. > However: > on the PowerBook the ata1 attaches as UDMA/100 ATA-6. > on the iBook I get UDMA/100 ATA-8 > > Is this ATA-8 all the reason of being faster? Is just the PB using a > slow HD? or it does not negotiate or get recognized to full speed? > Wondering! for both UDMA/100 indicates 100MBytes top speed. > ATA-8 appears to be a hybrid drive... with a cache, I don't think that > part can explain speed up for a 256MB file! internet search says it is > an 8MB cache. > > Riccardo > >
Re: kernel 5.9 no keyboard on PowerBook
On 12/9/20 3:22 PM, Riccardo Mottola wrote: > Also, good news, I updated today and again a 5.9 kernel was installed and > updated, > but it is 2020-11-27 and the keyboard works again! > (Interestingly, the external apple keyboard still doesn't work). Have you independently verified that the external Apple keyboard works correctly on this machine, i.e. does it work with MacOS or an older kernel version? Adrian -- .''`. John Paul Adrian Glaubitz : :' : Debian Developer - glaub...@debian.org `. `' Freie Universitaet Berlin - glaub...@physik.fu-berlin.de `-GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
Re: poor relative G4 performance iBook/PowerBook
Hi, Brian Morris wrote: % time cat huge_file_copy Install smartctl to check drive health I also have a PowerBook 1ghz which is quite a bit faster than my 1.5ghz tho could be because the former has a bit more ram and a newer hard drive and the latter has a defective 2nd ram slot I wanted to investigate this further - except smartctl, I did more extensive checks with file copying which are interesting, also, now that keyboard works again, I investigated better what dmesg says. Here some results: dd if=/dev/zero of=output.dat bs=1M count=128 PB: 158 MB/s iBook: 117 MB/s dd if=/dev/zero of=output.dat bs=1M count=256 PB: 9.4 28.5 MB/s iBook: 4.8s 56 MB/s Copy of 256MB file PB: 5.18 s iBook: 2.7 s This means that for 128MB there is some "cache effect" and there the PowerBook beats the iBook. But with 256MB the iBook is almost twice as fast! 28.5 MB/s is less than UDMA/33... quite slow... Both the iBook and PowerBook still have PATA of course, since no serial ata devices existed back then the iBook has [ 3.774751] pata-pci-macio 0002:20:0d.0: Activating pata-macio chipset UniNorth ATA-6, Apple bus ID 3 [ 3.775921] scsi host0: pata_macio [ 4.802731] pata-macio 0.0002:ata-3: Activating pata-macio chipset KeyLargo ATA-3, Apple bus ID 0 [ 4.810545] scsi host1: pata_macio I bet the first one is the Hard Disk, the second one is for the optical drive. The PowerBook appears also to have the same UniNorth and KeyLargo... so I doubt there is a driver issue? Apple reused the same chips. However: on the PowerBook the ata1 attaches as UDMA/100 ATA-6. on the iBook I get UDMA/100 ATA-8 Is this ATA-8 all the reason of being faster? Is just the PB using a slow HD? or it does not negotiate or get recognized to full speed? Wondering! for both UDMA/100 indicates 100MBytes top speed. ATA-8 appears to be a hybrid drive... with a cache, I don't think that part can explain speed up for a 256MB file! internet search says it is an 8MB cache. Riccardo
Re: kernel 5.9 no keyboard on PowerBook
Hi, Stan Johnson wrote: Adrian, Yes, the 5.9.0 kernel works on my PB G4 12". The built-in keyboard is detected as ADB. In a Google search, everything I was able to find indicated that G4 PowerBooks use ADB for their built-in keyboards and trackpads. I can say that 5.7 kernels recognize the internal stuff on usb, as opposed to my iBook. There were different revisions of the PowerBooks. I had a further strange test with 5.9 "not working" by pure change. I attached an Apple USB keyboard.. and it does not get recognized, but a generic "PC" keyboard works... Also, good news, I updated today and again a 5.9 kernel was installed and updated, but it is 2020-11-27 and the keyboard works again! (Interestingly, the external apple keyboard still doesn't work). I notice that on the USB keyboard the "caps lock" led doesn't light up when in console, but on my iBook it works (where it has ADB). Under X11 instead the led then lights up! Riccardo
Re: Updated installer images 2020-12-02
On 12/5/20 2:44 PM, Jeffrey Walton wrote: >> Your blank screen may have a different cause such as such as missing >> firmware for your graphics card which is a result of Debian Ports >> currently not having access to contrib and non-free during >> installation - something I can't influence but that's supposed to >> be fixed in the future. > > I think it is interesting (in a morbid sort of way) that a Debian > maintainer is not allowed access to the port to fix things. > > What is the problem? Who is putting barriers in your way? This problem involves changes to the FTP servers for Debian Ports to which only the FTP admins in Debian have access to. It's not uncommon that not everyone in a larger project has access to every resource. You have dedicated teams. So, in the end, I can just keep asking them and play the waiting game. Adrian -- .''`. John Paul Adrian Glaubitz : :' : Debian Developer - glaub...@debian.org `. `' Freie Universitaet Berlin - glaub...@physik.fu-berlin.de `-GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
